# Semiparametric detection of significant activation for brain fMRI

**Authors:** Chunming Zhang, Tao Yu

arXiv: 0808.0989 · 2008-08-08

## TL;DR

This paper introduces a semiparametric inference method for fMRI data analysis that improves the detection of activated brain regions over traditional linear models and existing tools.

## Contribution

It develops a semiparametric testing framework using local linear estimation for more accurate activation detection in fMRI data.

## Key findings

- Semiparametric tests follow chi-squared distributions under null hypotheses.
- The proposed method has higher detection efficiency than AFNI and FSL.
- Simulation and real data demonstrate improved activation detection.

## Abstract

Functional magnetic resonance imaging (fMRI) aims to locate activated regions in human brains when specific tasks are performed. The conventional tool for analyzing fMRI data applies some variant of the linear model, which is restrictive in modeling assumptions. To yield more accurate prediction of the time-course behavior of neuronal responses, the semiparametric inference for the underlying hemodynamic response function is developed to identify significantly activated voxels. Under mild regularity conditions, we demonstrate that a class of the proposed semiparametric test statistics, based on the local linear estimation technique, follow $\chi^2$ distributions under null hypotheses for a number of useful hypotheses. Furthermore, the asymptotic power functions of the constructed tests are derived under the fixed and contiguous alternatives. Simulation evaluations and real fMRI data application suggest that the semiparametric inference procedure provides more efficient detection of activated brain areas than the popular imaging analysis tools AFNI and FSL.

## Full text

Tomesphere does not host the full text of arXiv papers. Read it on arXiv: https://arxiv.org/abs/0808.0989

---
Source: https://tomesphere.com/paper/0808.0989